Analyzing GeoMx-NGS Protein Expression Data with GeomxTools

Filter ROIs with low signal

First, we filtered 7 ROIs with low signals of housekeeping genes which geometric mean are lower than 4.

Then we also calculated the geometric mean of IgG signals in each ROI.

According to the correlation between housekeeping and IgG background signal are concordant with each other.

Filter proteins with low signal

We computed the signal-to-background ratio per target, which is simply each ROI’s data divided by its IgG geomean. And we removed 26 targets with lower signals than the control group in the ROIs.

Data normalization (Housekeeping proteins)

We finally chose the housekeeping normalization for data normalization.

PCA of normalized data

After the data normalization and the low targets removal, we tested the correlation between each ROIs.

Differential signals comparison

We performed the comparisons in M12, F12, M3, and F3, within the Hip and Non_hip respectively. L vs. C LPE vs. L LPG vs. L LPG vs. LPE